The establishment of a chronic asynchronous heart failure (HF) model by rapid pacing combined with left bundle branch ablation is presented. Two-dimensional speckle tracking imaging and aortic velocity time integral are applied to validate this stable HF model with left ventricular asynchrony and the benefits of cardiac resynchronization therapy.
It is now well recognized that heart failure (HF) patients with left bundle branch block (LBBB) derive substantial clinical benefits from cardiac resynchronization therapy (CRT), and LBBB has become one of the important predictors for CRT response. The conventional tachypacing-induced HF model has several major limitations, including absence of stable LBBB and rapid reversal of left ventricular (LV) dysfunction after cessation of pacing. Hence, it is essential to establish an optimal model of chronic HF with isolated LBBB for studying CRT benefits. In the present study, a canine model of asynchronous HF induced by left bundle branch (LBB) ablation and 4 weeks of rapid right ventricular (RV) pacing is established. The RV and right atrial (RA) pacing electrodes via the jugular vein approach, together with an epicardial LV pacing electrode, were implanted for CRT performance. Presented here are the detailed protocols of radiofrequency (RF) catheter ablation, pacing leads implantation, and rapid pacing strategy. Intracardiac and surface electrograms during operation were also provided for a better understanding of LBB ablation. Two-dimensional speckle tracking imaging and aortic velocity time integral (aVTI) were acquired to validate the chronic stable HF model with LV asynchrony and CRT benefits. By coordinating ventricular activation and contraction, CRT uniformed the LV mechanical work and restored LV pump function, which was followed by reversal of LV dilation. Moreover, the histopathological study revealed a significant restoration of cardiomyocyte diameter and collagen volume fraction (CVF) after CRT performance, indicating a histologic and cellular reverse remodeling elicited by CRT. In this report, we described a feasible and valid method to develop a chronic asynchronous HF model, which was suitable for studying structural and biologic reverse remodeling following CRT.
Advanced chronic HF is a leading cause of mortality for various cardiovascular diseases. A subset of patients with congestive heart failure (CHF) also develop ventricular conduction discoordination that aggravates symptoms and prognosis. CRT, also referred to as biventricular pacing, has been introduced as an alternative therapy for these patients for over 20 years1,2. Unfortunately, about 20-40% of the patients show poor response to CRT. Since then, many studies have been carried out in order to maximize CRT response3. It is now well recognized that patients with LBBB could benefit more from CRT than those with non-LBBB4, since an LBBB pattern causes a larger magnitude of cardiac dyssynchrony due to asymmetry in the freedom of wall movement between septal and lateral walls. Meanwhile recent studies have begun exploring changes in gene expression and molecular remodeling associated with CRT5. Accompanying the structural reverse remodeling induced by CRT, cellular and molecular reversion to a normal level is of great interest6. Hence, it is essential to establish an optimal model of CHF with isolated LBBB for studying CRT benefits.
Chronic, rapid ventricular pacing was once used to produce CHF in a canine model. RV pacing could undoubtedly produce delayed LV contraction as a model of the LBBB-like contraction pattern. However, this type of functional asynchrony with an intact conduction system may not emulate anatomical LBBB and is not considered an appropriate model for studying CRT performance, the essence of which is to coordinate impaired electrical activation and myocardial contraction. Rapid restoration of LV contractility and partial recovery of LV dimensions after cessation of pacing were also reported7.
Experimental studies have induced chronic LBBB by RF ablation to establish asynchronous ventricular contraction8. A combination of reduction in global pump function and regional invalid mechanical work could exacerbate CHF by generating cardiac inefficiency as well as cardiac remodeling at the tissue, cellular, and molecular levels. In LBBB hearts, workload is lowest in the septum and highest in the LV lateral wall. As a consequence, cardiac remodeling is most pronounced in the lateral wall9. The purpose of the present study is: (i) to advance a stable and chronic HF model with interventricular and intraventricular mechanical asynchrony by means of rapid RV pacing in combination with LBB ablation; (ii) to confirm dyssynchronous HF in our model and CRT benefits in coordinating contraction by two-dimensional speckle tracking echocardiography and aVTI; and (iii) to preliminarily explore cellular reverse remodeling elicited by CRT.
Fifteen male beagle dogs (12 to 18 months old, weighing around 10.0-12.0 kg) were purchased and subjected to experiments. All procedures were performed in compliance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (Publication No. 85-23, revised 1996) and were approved by the Animal Care Committee in Zhongshan Hospital, Fudan University. Figure 1 shows the schematic workflow for all protocol steps.
1. Pre-surgery Preparation and Baseline Data Collection
2. Epicardial LV Pacing Electrode Implantation
3. RA and RV Pacing Electrodes Implantation
4. LBB Ablation
5. Rapid Pacing for HF Induction
6. Cardiac Resynchronization Therapy Performance
7. Sacrifice Animals and Histological Analysis
Successful LBB Ablation:
Figure 2 represents a typical surface and intracardiac electrogram in the course of catheter ablation. The mean LBP-V measured is 18.8 ±2.8 ms, which was about 10 ms shorter than the baseline H-V interval (28.8 ±2.6 ms, p <0.01). The QRS duration prolonged from 59.2 ±6.8 ms to 94.2 ±8.6 ms (p <0.01) after LBB ablation. The loss of the LBP electrogram confirmed successful LBB ablation.
A Chronic Dyssynchronous CHF Model and CRT Benefits Quantified by Echocardiography:
Baseline echocardiographic parameters showed no significant difference among the sham, HF control, and CRT groups. As was published in our previous data10, an obvious deteriorated cardiac function characterized by increased LVEDV and LVESV, and decreased LVEF could be observed in the HF control group at the end of the experiment (Figure 3). CRT improved cardiac function with decreased LVEDV and LVESV, and increased LVEF. For speckle tracking analysis, tri-plane apical longitudinal views including A4C, A2C, and APLAX were acquired simultaneously. After tracing each apical view, the longitudinal strain curves of six segments from each plane were obtained. Then the TTP and PSD were calculated. As a result, an increased asynchrony index (PSD) was induced in the HF control group compared with the sham group (51.6 ±5.9 ms vs. 32.6 ±2.3 ms, p <0.01); while CRT corrected LV asynchrony, as exhibited by a significantly lower PSD (44.0 ±4.6 ms vs. 51.6 ±5.9 ms, p <0.05). Furthermore, the HF control animals presented a significantly lower aVTI than the sham group (8.09 ±1.19 cm vs. 14.53 ±2.38 cm, p <0.01), which was significantly increased in the CRT group (10.92 ±1.31 cm vs. 8.09 ±1.19 cm, p <0.05) (Figure 3 and Figure 4).
Histologic and Cellular Reverse Remodeling Induced by CRT:
Myocardial tissues excised from the LV lateral wall were subjected to histologic analysis. Compared with the sham group, a remarkably decreased cardiomyocyte diameter was noted in the HF control group (4.77 ±0.86 µm vs. 7.68 ±1.25 µm, p <0.01), which might be responsible for the LV dilation. Masson trichrome staining revealed a significant increase of CVF in the HF control group in contrast with the sham group (12.56 ±2.10% vs. 1.88 ±0.23%, p <0.01). However, 8 weeks of CRT performance resulted in a significant restoration of cardiomyocyte diameter (6.26 ±0.93 µm vs. 4.77 ±0.86 µm, p <0.01) and CVF (6.28 ±1.61% vs. 12.56 ±2.10%, p <0.01) compared with the HF control group, indicating a biologic reverse remodeling invoked by CRT (Figure 5).
Figure 1: The schematic workflow of all protocol steps. Please click here to view a larger version of this figure.
Figure 2: 12-lead ECG and intracardiac electrogram recorded before (A) and after (B) catheter ablation. (A) Typical surface and intracardiac electrogram at a successful ablation site. Right-sided His bundle potential was mapped by the distal electrode of the quadripolar catheter with an H-V interval of 28 ms. The LBP was mapped by the ablation catheter with an LBP-V interval of 17 ms. The LBP-V interval was 11 ms shorter than the H-V interval. (B) Typical LBBB morphology after successful ablation. The QRS duration prolonged from 63 ms to 95 ms after LBB ablation, which was positive in leads I, aVF, V6, with notched R wave, and negative in lead V1. The LBP disappeared and the right-sided His bundle potential still existed after ablation. Please click here to view a larger version of this figure.
Figure 3: Bar graphs expressed as mean ± SD for LVEDV, LVESV, LVEF, PSD, and aVTI among the three experimental groups (n = 5 for each) at baseline and the end of experiment, respectively. Values between the experimental groups were compared using one-way ANOVA test. Compared with the sham group, *p <0.05, **p <0.01; Compared with the HF control group, #p <0.05, ##p <0.01. Please click here to view a larger version of this figure.
Figure 4: Speckle tracking strain imaging and aortic velocity time integral measurement. (A) Two-dimensional longitudinal strain analysis using speckle tracking imaging from 3 standard apical views. A1 showed tri-plane apical longitudinal views acquired using 4VD transducer of GE VIVID E9. Images were carefully adjusted to ensure that apical four chamber view (A4C), two chamber view (A2C), and long axis view (APLAX) were displayed at the same time. A2 displayed an example of longitudinal strain curves of six segments created by a tracking algorithm from APLAX view. Segments of basal-posterior wall, mid-posterior wall, apical-posterior wall, basal-anterior septum, mid-anterior septum and apical-anterior septum were automatically defined. A3 showed the time to peak longitudinal strain (TTP) of each segment calculated with QRS onset as a reference when all the segmental time-strain curves were constructed from the three apical views. A significantly higher dispersion of TTP could be observed in the HF control group, which was formulated as standard deviation. CRT performance significantly reduced the difference between TTP of each segment. (B) Assessment of aortic velocity time integral averaged from 3 consecutive beats. B1, B2, and B3 represent typical images of the sham group, HF control group, and CRT group, respectively. Please click here to view a larger version of this figure.
Figure 5: Typical photograph of the HE staining (400X) and Masson's trichrome staining (400X). Diameters of myocardial fibers were measured from longitudinally cut sections, and collagen volume fraction (CVF) was assessed from the percentage of fibrotic area divided by total tissue area. Scale bars = 50 µm. Please click here to view a larger version of this figure.
Dilated cardiomyopathy constitutes a major cause of CHF, which is characterized by ventricular dilation, systolic dysfunction with reduced LVEF, and abnormalities of diastolic filling11. Since chronic tachycardia-mediated HF is a recognized clinical condition, rapid pacing of either atrium or ventricle for at least 3 to 4 weeks serves as a frequently used animal model to induce CHF11. Hemodynamic changes occur as soon as 24 h after rapid pacing, with continued deterioration of cardiac function for up to 3 to 5 weeks. However, the recovery from pacing-induced HF is a dramatic and unique feature of this model, accompanied by a reversal of neurohormonal activation, indicating a reversible nature of this myopathy. It is documented that LVEF shows significant recovery within 1 to 2 weeks after termination of pacing and nearly all hemodynamic variables return towards normal levels at 4 weeks after cessation of rapid pacing12. Hence, the prevention of cardiac function recovery on the discontinuation of pacing is of great importance in this attractive model.
LBBB could result in delayed LV activation and a corresponding delayed LV systole. Asynchronous contraction of the septum and the LV free wall performs a disproportionate amount of net myocardial work, and the work is wasted in both regions. Though a mere LBBB produces a low-grade myopathy, the synergy between HF and LBBB may produce substantial functional and clinical decline over time, which could be ameliorated by CRT. The functional LBBB induced by RV pacing is temporary, which is very different from the case where an anatomical LBBB is present. In the present study, a permanent LBBB was created by catheter ablation and its presence was confirmed during the subsequent experiments. Canines have a relatively longer, more left side oriented penetrating bundle of His and common left bundle, which may account for the high success rate of LBB ablation. LBP is located between His bundle and Purkinje potentials. Correct identification of LBP and a guarantee of A:V electrogram ratio <1:10 favors successful LBB ablation and avoidance of complete A-V block13. The common left bundle is divided into anterior and posterior fascicles at the proximal one-thirds along the muscular ventricular septum. If the ablation catheter is positioned on a distal portion of the bundle branch, an anterior or posterior fascicle might be ablated. However, the ablation of these fascicles could not obviously prolong the QRS duration. Based on a previous study, the QRS duration could increase by 40-50 ms following LBB ablation13. In the present study, the QRS prolongation averaged 35 ms, which might be due to different animal species. On intracardiac electrogram, the mean LBP-V interval for successful ablation measured about 16-19 ms, usually 10 ms shorter than the H-V interval, neither too close nor too far from the His bundle. In addition, the LBP usually disappeared after successful ablation14.
A previous study has reported that rapid pacing for at least 3 to 4 weeks produces a reliable and reproducible HF model11. There may exist some difference among different animals as for the necessary period of tachypacing. So, echocardiography was performed every 2 weeks during rapid pacing. None of the animals showed an LVEF <35% after 2 weeks of tachypacing, suggesting that 3 to 4 weeks of rapid pacing is essential. After 4 weeks, once the LVEF was below 35%, rapid pacing was terminated. Such a strategy helped to uniform the baseline HF severity. In addition, since RV apical (RVA) pacing has long been proven to induce LV dyssynchrony and HF15, we selected RVA instead of RA for rapid pacing. Thus, rapid pacing-induced HF with superimposed LBBB-induced dyssynchrony in our study helped to establish a model of stable and chronic dyssynchronous HF. More importantly, the LV systolic dysfunction hardly recovered in up to 8 weeks of observation in the control group. Such an animal model favored investigation of CRT benefits instead of self-recovery.
To establish the HF model, first we implanted the LV epicardial lead via left thoracotomy. After 2 weeks of recovery from thoracotomy, we implanted the RV and RA leads via a jugular vein approach, followed by LBB ablation. Although limited left thoracotomy, muscle sparing, and rib preservation strategies are excellent minimally invasive approaches for exposure of the LV lateral wall, operative trauma and postoperative infection are still associated with high mortality. So, the LV lead implantation was performed before other procedures. Only those surviving 2 weeks after the operation are submitted to LBB ablation and rapid pacing. Overall, this was an economical strategy.
Echocardiographic data demonstrated persistence of significant systolic dysfunction, increased ventricular volumes, and higher asynchrony index in our CHF model. CRT improved cardiac function with reduced asynchrony index. Speckle tracking strain analysis is a novel method which permits the assessment of myocardial deformation. It has proven to be significantly associated with long-term outcome after CRT and has additive prognostic value to routine selection criteria for CRT. Of the three different patterns of myocardial deformation including radial strain, circumferential strain, and longitudinal strain, it is still under debate with conflicting data, which one used for LV dyssynchrony index may best predict CRT response16,17. However, it is reported that global longitudinal strain consistently showed good reproducibility, while reproducibility was moderate for circumferential strain and poor in the radial direction18. Therefore, in the present study, we adopted the apical tri-plane longitudinal strain analysis as the LV asynchrony index by calculating PSD. A higher PSD indicated a severer asynchrony. aVTI has been commonly used for AV and VV delay optimization in CRT patients. Changes in aVTI can serve as a surrogate for changes in stroke volume as it is directly proportional to the LV outflow tract VTI19. Hence, we assessed aVTI to evaluate hemodynamic benefits from CRT. A higher aVTI suggested better LV systolic performance.
Cardiac fibrosis, as characterized by interstitial collagen and extracellular matrix deposit, is a hallmark of end-stage CHF. Recent studies have demonstrated that LV reverse remodeling after CRT is independently associated with diffuse interstitial myocardial fibrosis, which is assessed with myocardial T1 mapping cardiac magnetic resonance (CMR)20. Besides, CRT-induced LV reverse remodeling is also associated with a decreased plasma level of pro-fibrotic cytokines such as transforming growth factor (TGF)-β1 and osteopontin (OPN)21,22. In the present study, histologic analysis revealed decreased cardiomyocyte diameter and increased myocardial fibrosis in the failing heart at 8 weeks after cessation of rapid pacing, suggesting a histologic and cellular remodeling in our HF model. Along with the structural reverse remodeling, however, CRT restored myocyte configuration and alleviated collagen deposits. Such a histologic reverse remodeling yields more beneficial effects beyond CRT itself.
Recent recommendations for CRT implantation include persistent HF symptoms, impaired LV systolic function with LVEF ≤35%, LBBB QRS morphology, and a widened QRS duration4. Our experimental model is a practicable, reproducible, and stable HF model, which satisfies almost all these criteria. While it is noteworthy that our work established a canine model of non-ischemic dilated cardiomyopathy, it may not apply to other conditions such as valvular heart disease, congenital heart disease, ischemic HF, etc. Especially, coronary ligation or microembolization is commonly used to produce ischemic HF, which has a higher risk of sudden cardiac death. However, due to the discrepancy of myocardial scar burden in ischemic HF, it is not easy to objectively evaluate CRT benefits. By contrast, our experimental model is relatively homogeneous and is a suitable model for studying CRT performance, including electrical behavior, echocardiographic assessment, and biologic and molecular modifications.
The authors have nothing to disclose.
This work is funded by National Natural Science Foundation of China (81671685) and Shanghai Commission of Health and Family Planning (No. 201440538)
Closed iv catheter system (0.9mm×25mm) | Becton Dickinson Medical | 5264442 | Used as venous retention needle |
Sodium pentobarbital | Sigma-Aldrich Company | 130205 | For anesthesia |
Pet clipper | Wuhan Shernbao pet supplies Co., Ltd. | PGC-660 | For hair shaving |
Electrocardiograph | Shanghai photoelectric medical electronic instrument Co., Ltd. | ECG-6511 | For electrocardiogram recording |
Echocardiograph | GE-Vingmed Ultrasound Company | VIVID E9 | For echocardiographic assessment |
EchoPAC software | GE healthcare | Version201 | Offline analysis |
Laryngoscope | Shanghai Medical Instrument Co., Ltd | Orotracheal intubation | |
Endotracheal tube | SIMS Portex Inc, UK | 274093 | Orotracheal intubation |
Volume cycled respirator | Newport Corporation | C100 | Artificial ventilation |
HeartStart XL Defibrillator/Monitor | Philips Medical Systems | M4735A | Electrocardiogram monitor during operation |
Benzalkonium Bromide Tincture | Shanghai Yunjia Pharmaceutical Co., Ltd. | H31022694 | Used for skin disinfection |
Rib retractor | Shanghai Medical Instrument Co., Ltd. | For thoracotomy | |
4-0 suture | Shanghai Pudong Jinhuan Medical Products Co., LTD. | 24L1005 | Suture of LV epicardial electrode |
2-0/T suture | Shanghai Pudong Jinhuan Medical Products Co., LTD. | 11M0505 | Suture of pacing leads, fascia, vessels, etc. |
0-suture | Shanghai Pudong Jinhuan Medical Products Co., LTD. | 11P0501 | Skin suture |
penicillin powder | North China Pharmaceutical Co., Ltd. | F6034105 | |
DSA X-ray machine | Philips | Allura Xper FD10 | X-ray for fluoroscopy |
LV pacing electrode | Medtronic, Inc. | LBT 4965 | |
RV pacing electrode | St. Jude Medical | Tendril 1888 | |
RA pacing electrode | St. Jude Medical | IsoFlex 1642T | |
Pacemaker pulse generator | Medtronic, Inc. | Enpulse E2DR01 | For rapid RV pacing |
CRT pulse generator | St. Jude Medical | Anthem PM 3212 | For CRT performance |
Multi-channel electrophysiologic recorder | GE Medical Systems | 2003232-004 | For surface and intracardiac electrogram |
Catheter input module | GE Medical Systems | 301-00202-08 | Multiple pole switches for stimulation or recording |
Radiofrequency generator | Johnson-Johnson Company | ST-4460 | For RF current delivery |
Cordless return electrode | Covidien | E7509 | For current circuit formation |
Cordis 6-Fr sheath | Johnson-Johnson Company | 504-606X | Access for mapping catheter |
Cordis 7-Fr sheath | Johnson-Johnson Company | 504-607X | Access for mapping and ablation catheter |
6-Fr quadripolar catheter | Johnson-Johnson Company | F6QRA005RT | Mapping catheter |
7-Fr 4mm-tip steerable ablation catheter | St. Jude Medical | 402823 | Mapping and ablation catheter |
Prucka Cardio-Lab®2000 | GE Medical Systems | 6.9.00.000 | Software package for electrogram recording |
Heparin | Haitong Pharmaceutical Co., Ltd | 160505 | Anticoagulant during catheter ablation |
Digital image analysis system | Leica Microsystems | Qwin V3 | For histologic analysis |